Evaluating Adversarial Robustness of Concept Representations in Sparse Autoencoders.

Aaron J. Li, Suraj Srinivas, Usha Bhalla, Himabindu Lakkaraju

Browse the full EACL paper archive.